A Newton inexact interior-point method for large scale nonlinear optimization problems

In this paper, we describe a variant of the Newton Interior-Point method in [8] for nonlinear programming problems. In this scheme, the perturbation parameter can be chosen within a range of, values and we can use an iterative method for approximately solving the reduced linear system arising at each step. We have devised the inner termination rule which guarantees the global convergence of this Newton Inexact Interior-Point method. We remark that the required assumptions are weaker than those stated in [8], as shown by some numerical examples. This research was supported by the Italian Ministry for Education, University and Research (MIUR), FIRB Project No. RBAU01JYPN.     

Path-following barrier and penalty methods for linearly constrained problems

In the present paper some barrier and penalty methods (e.g. logarithmic barriers, SUMT, exponential penalties), which define a continuously differentiable primal and dual path, applied to linearly constrained convex problems are studied, in particular, the radius of convergence of Newton’s method depending on the barrier and penalty para-meter is estimated, Unlike using self-concordance properties the convergence bounds are derived by direct estimations of the solutions of the Newton equations. The obtained results establish parameter selection rules which guarantee the overall convergence of the considered barrier and penalty techniques with only a finite number of Newton steps at each parameter level. Moreover, the obtained estimates support scaling method which uses approximate dual multipliers as available in barrier and penalty methods     

An unconstrained convex programming view of linear programming

The major interest of this paper is to show that, at least in theory, a pair of primal and dual -optimal solutions to a general linear program in Karmarkar's standard form can be obtained by solving an unconstrained convex program. Hence unconstrained convex optimization methods are suggested to be carefully reviewed for this purpose.     

An inexact approach to solving linear semi-infinite programming problems

In this paper, we propose an “inexact solution” approach to deal with linear semi-infinite programming problems with finitely many variables and infinitely many constraints over a compact metric space. A general convergence proof with some numerical examples are given and the advantages of using this approach are discussed     

Asymptotic analysis of a path-following barrier method for linearly constrained convex problems *

In the present paper the logarithmic barrier method applied to the linearly constrained convex optimization problems is studied from the view point of classical path-following algorithms. In particular, the radius of convergence of Newton's method which depends on the barrier parameter itself is estimated in standard norms, being independent of the parameter, without explicitly using self-concordance properties. The obtained results establish a parameter selection rule which guarantees the overall convergence of a barrier technique with only one Newton step at each parameter level and the complexity of the method can be estimated.     

On solving convex quadratic semi-infinite programming probelms

In this paper we study a from of convex quadratic semi-infinite programming problems with finitely many variables and infinitely many constraints over a compact metric space. An entropic path-following algorithum is introduced with a convergence proof. Some practical implementations and numerical experiments are also included     

A Barrier Function Method for the Nonconvex Quadratic Programming Problem with Box Constraints

In this paper a barrier function method is proposed for approximating a solution of the nonconvex quadratic programming problem with box constraints. The method attempts to produce a solution of good quality by following a path as the barrier parameter decreases from a sufficiently large positive number. For a given value of the barrier parameter, the method searches for a minimum point of the barrier function in a descent direction, which has a desired property that the box constraints are always satisfied automatically if the step length is a number between zero and one. When all the diagonal entries of the objective function are negative, the method converges to at least a local minimum point of the problem if it yields a local minimum point of the barrier function for a sequence of decreasing values of the barrier parameter with zero limit. Numerical results show that the method always generates a global or near global minimum point as the barrier parameter decreases at a sufficiently slow pace.     

On the entropic regularization method for solving min-max problems with applications

Consider a min-max problem in the form of min _xX max_1im {f _i (x)}. It is well-known that the non-differentiability of the max functionF(x) max_1im {f _i (x)} presents difficulty in finding an optimal solution. An entropic regularization procedure provides a smooth approximationF _p(x) that uniformly converges toF(x) overX with a difference bounded by ln(m)/p, forp > 0. In this way, withp being sufficiently large, minimizing the smooth functionF _p(x) overX provides a very accurate solution to the min-max problem. The same procedure can be applied to solve systems of inequalities, linear programming problems, and constrained min-max problems.This research work was supported in part by the 1995 NCSC-Cray Research Grant and the National Textile Center Research Grant S95-2.     

Solving staircase linear programs by the simplex method, 2: Pricing

This and a companion paper consider how current implementations of the simplex method may be adapted to better solve linear programs that have a staged, or ‘staircase’, structure. The preceding paper considered ‘inversion’ routines that factorize the basis and solve linear systems. The present paper examines ‘pricing’ routines that compute reduced costs for nonbasic variables and that select a variable to enter the basis at each iteration. Both papers describe extensive (although preliminary) computer experiments, and can point to some quite promising results. For pricing in particular, staircase computation strategies appear to offer modest but consistent savings; staircase selection strategies, properly chosen, may offer substantial savings in number of iterations, time per iteration, or both.     

Solving staircase linear programs by the simplex method, 1: Inversion

This and a companion paper consider how current implementations of the simplex method may be adapted to better solve linear programs that have a staged, or staircase, structure. The present paper looks at inversion routines within the simplex method, particularly those for sparse triangular factorization of a basis by Gaussian elimination and for solution of triangular linear systems. The succeeding paper examines pricing routines. Both papers describe extensive (though preliminary) computational experience, and can point to some quite promising results.     

A class of methods for linear programming

A class of methods is presented for solving standard linear programming problems. Like the simplex method, these methods move from one feasible solution to another at each iteration, improving the objective function as they go. Each such feasible solution is also associated with a basis. However, this feasible solution need not be an extreme point and the basic solution corresponding to the associated basis need not be feasible. Nevertheless, an optimal solution, if one exists, is found in a finite number of iterations (under nondegeneracy). An important example of a method in the class is the reduced gradient method with a slight modification regarding selection of the entering variable.     

模糊线性规划问题的一种新的单纯形算法

提出求解模糊线性规划问题的一种新的思路 ,就是应用单纯形法先求解与 (FLP)相应的普通线性规划问题 ,通过模糊约束集与模糊目标集的隶属度的比较 ,获得两个集合交集的最优隶属度 ,将此最优隶属度代入最优单纯形表中 ,即可求得 (FLP)的解。本算法只需在一张适当的迭代表台上执行单纯形迭代过程 ,简捷方便适用     

Large-scale linearly constrained optimization

An algorithm for solving large-scale nonlinear programs with linear constraints is presented. The method combines efficient sparse-matrix techniques as in the revised simplex method with stable quasi-Newton methods for handling the nonlinearities. A general-purpose production code (MINOS) is described, along with computational experience on a wide variety of problems.This research was supported by the U.S. Office of Naval Research (Contract N00014-75-C-0267), the National Science Foundation (Grants MCS71-03341 A04, DCR75-04544), the U.S. Energy Research and Development Administration (Contract E(04-3)-326 PA #18), the Victoria University of Wellington, New Zealand, and the Department of Scientific and Industrial Research Wellington, New Zealand.     

Scaling,shifting and weighting in interior-point methods

We examine certain questions related to the choice of scaling, shifting and weighting strategies for interior-point methods for linear programming. One theme is the desire to make trajectories to be followed by algorithms into straight lines if possible to encourage fast convergence. While interior-point methods in general follow curves, this occurrence of straight lines seems appropriate to honor George Dantzig's contributions to linear programming, since his simplex method can be seen as following either a piecewise-linear path inn-space or a straight line inm-space (the simplex interpretation).Dedicated to Professor George B. Dantzig on the occasion of his eightieth birthday.Research supported in part by NSF, AFOSR, and ONR through NSF Grant DMS-8920550.     

Quadratic interior-point methods in statistical disclosure control

The safe dissemination of statistical tabular data is one of the main concerns of National Statistical Institutes (NSIs). Although each cell of the tables is made up of the aggregated information of several individuals, the statistical confidentiality can be violated. NSIs must guarantee that no individual information can be derived from the released tables. One widely used type of methods to reduce the disclosure risk is based on the perturbation of the cell values. We consider a new controlled perturbation method which, given a set of tables to be protected, finds the closest safe ones - thus reducing the information loss while preserving confidentiality. This approach means solving a quadratic optimization problem with a much larger number of variables than constraints. Real instances can provide problems with millions of variables. We show that interior-point methods are an effective choice for that model, and, also, that specialized algorithms which exploit the problem structure can be faster than state-of-the art general solvers. Computational results are presented for instances of up to 1000000 variables.AMS Subject Classification: 90C06, 90C20, 90C51, 90C90Jordi Castro: Partially supported by the EU IST-2000-25069 CASC project and by the Spanish MCyT project TIC2003-00997.     

On the non-polynomiality of the relaxation method for systems of linear inequalities

A class of linear programs is given in which the relaxation method for inequalities, under the same operating rules as Khacian's method, is not polynomial in the length of the input. This result holds for any value of the relaxation parameter.This research was supported in part by the D.G.E.S. (Quebec), the N.S.E.R.C. of Canada under grant A 4152, and the S.S.H.R.C. of Canada.     

Parabolic target space and primal-dual interior-point methods

In this paper we develop new primal-dual interior-point methods for linear programming problems, which are based on the concept of parabolic target space. We show that such schemes work in the infinity-neighborhood of the primal-dual central path. Nevertheless, these methods possess the best known complexity estimate. We demonstrate that the adaptive-step path-following strategies can be naturally incorporated in such schemes.     

Nonlinear programming methods in the presence of noise

The problem of minimizing a nonlinear function with nonlinear constraints when the values of the objective, the constraints and their gradients have errors, is studied. This noise may be due to the stochastic nature of the problem or to numerical error.Various previously proposed methods are reviewed. Generally, the minimization algorithms involve methods of subgradient optimization, with the constraints introduced through penalty, Lagrange, or extended Lagrange functions. Probabilistic convergence theorems are obtained. Finally, an algorithm to solve the general convex (nondifferentiable) programming problem with noise is proposed.Originally written for presentation at the 1976 Budapest Symposium on Mathematical Programming.     

A practicable way for computing the directional derivative of the optimal value function in convex programming

Formulas for computing the directional derivative of the optimal value function or of lower or upper bounds of it are well-known from literature. Because they have as a rule a minmax structure, methods from nondifferentiable optimization are required.

Considering a fully parametrized convex problem, in the paper the mentioned minmax formulas are transformed into usual programming problems. Although they are nonconvex in general, the computational effort is much lower than that for minmax problems. In several special cases, for instance, for linear least squares problems, linear programming problems arise.     

A globally convergent constrained quasi-Newton method with an augmented lagrangian type penalty function

The recently proposed quasi-Newton method for constrained optimization has very attractive local convergence properties. To force global convergnce of the method, a descent method which uses Zangwill's penalty function and an exact line search has been proposed by Han. In this paper a new method which adopts a differentiable penalty function and an approximate line is presented. The proposed penalty function has the form of the augmented Lagrangian function. An algorithm for updating parameters which appear in the penalty function is described. Global convergence of the given method is proved.